Sampling error
Eg: can be caused due to random
variations in the estimate.
-
The expected value of such variation = zero
-
In the homogeneous population there will be
lesser chance for random variation.
o So sampling
error = smaller magnitude.
-
As sample size = sampling error /////
-
But larger sample result in more systematic bias
and it leads more expenditure in data collection.
-
If sampling = complex greater chance = error
It good sample will save time / money
/ human rexureey
Sampling can be classified into 3:
1 probably sampling
2 quasi probably
sampling
3 non-probably sampling
probably sampling
in
this method each unit of the population has an equal chance of selection.
This = 3 type
1 simple
random sampling
2 stratified
random sampling
3 cluster
sampling
1 simple random sampling.
In
simple random every element of the population has equal chande of selection in
the sampl.
It
is used when complete and accurate sampling is available 1 lottery method 2
random number taste 3 computer program method etc. can be used to draw sample
randomly.
There
are used to avoid bias.
Eg:
of /////
2 stratified random sampling
In
this population divided in to “strata” and then random sample is used from each
strata.
Eg:
people visiting in mall then can be categorised in to different strata on the
basis of age / gender / companionate, then random sample can be chosen from
each strata.
3 cluster sampling
The
clusters will be formed from the population and then a cluster is chosen
randomly.
Difference between strata &
cluster =
Strata:
homogeneous (dosha quality ulla ) group of population.
Cluster:
heterogeneous
groups.
Sample
is chosen from each strata.
When
a cluster is selected all it’s unit are selected.
Quasi probability sampling
2 types:
1 systematic
sampling
2 multistage
sampling
systematic
sampling:
1st person chosen
randomly then subsequent person chosen by next N th unit of the position.
4th / 14th
/ 24th / 34th….etc.
Multi-stage
sampling
The sample is chosen in multi stage.
Eg:
there are total 190 college in 19 districts of Punjab.
Such:
each district = 10 colleges.
1000
student study in each college.
A
study is conducted which requires collection of data from total 1000 students
of 10 college & 5 district.
So
researcher choose 5/19 & 2college/district = 5*2=10
1000/10
student 100 students from each 10 colleges.
Non-probability sampling
Selection
is not random in such sampling.
Quota
sampling:
It
is similar like stratified random sampling.
In
quota sampling the population is divided into groups then each group is
allocated. Quota of same proportion has it in the total population.
Suppose:
in the population of 500 researchers
50%
science
40%
social science
10%
language
So
select sample 100/500
in 50% researchers from
science
40% social science
10% language
Purposive sampling (judgement sampling)
Sample
is selected as per the judgement of the researchers.
Eg:
selecting person who can provide information about research.
So
those unit of population are selected who meets the purpose of study.
It
used when population = small.
Such
sample may not be representative sample.
Samskrithi's Teaching Aptitude Notes Previous Pages and Next Page Click Below Links
For RA Part -06 Notes - Click Here
Research Aptitude Notes Part -08- Click Here
Post a Comment
Post a Comment